Bárbara Costa1,2,3,4, Co-supervisor / Second author Maria João Gouveia4,5, Supervisor / Last author Nuno Vale1,2,3
1PerMed Research Group, Center for Health Technology and Services Research (CINTESIS), Porto, Portugal, 2Faculty of Medicine, University of Porto, 3Department of Community Medicine, Health Information and Decision (MEDCIDS), Faculty of Medicine, University of Porto, 4Centre for Parasite Biology and Immunology, National Health Institute Dr. Ricardo Jorge, Porto, Portugal; , 5Center for Study in Animal Science of University of Porto (CECA-ICETA UP)
Introduction: Pregnant women with chronic conditions such as human immunodeficiency virus (HIV) face unique challenges in medication management due to the physiological changes of pregnancy [1]. These changes, including alterations in drug metabolism, absorption, and elimination, complicate dosing and pharmacokinetic (PK) prediction [2]. Antiretroviral (ARV) therapy is critical for maintaining virologic suppression in HIV patients and preventing perinatal transmission [3]. Physiologically based pharmacokinetic (PBPK) and population pharmacokinetic (PopPK) modeling of ARVs have made substantial progress, enhancing our understanding of drug behavior in this complex patient population while accounting for cytochrome P450 (CYP450) enzymes’ contribution to drug-drug interactions (DDIs) [4,5]. The management of comorbidities like epilepsy further complicates treatment in this population [6]. Lamotrigine (LTG), a common antiepileptic, and efavirenz (EFV), a key ARV agent, both undergo significant PK changes during pregnancy. LTG clearance increases markedly due to enhanced glucuronidation, often requiring dose adjustments. The enzyme responsible for this process is UGT2B7, with a notable contribution from UGT1A4 as well [7]. EFV metabolism is affected by CYP2B6 polymorphisms, complicating its management in pregnancy. Though not primarily metabolized by UGT1A1, EFV may influence the metabolism of UGT1A1 substrates [9]. Significant progress has been made in modeling individual drug behaviors. However, studies on drug-drug interactions, especially those involving UGT metablism, remain less common [10]. The interaction between LTG and EFV during pregnancy is not well-established, and there is still no clear consensus on dose adjustments for their co-administration. Therefore, understanding potential interactions is crucial for ensuring the safety of both the mother and the fetus, as well as maintaining the therapeutic efficacy of both medications Objective: This study aimed to develop PBPK models to quantify the interaction between LTG and EFV in pregnant patients, incorporating pregnancy-induced metabolic changes. Specifically, it emphasized the shared role of UGT1A1 in the metabolism of both drugs to optimize drug dosing through in silico modeling. Methods: This study employed a systematic approach to collect relevant PK data from peer-reviewed clinical studies focusing on both pregnant and non-pregnant individuals treated with LTG and EFV. The data collection prioritized parameters such as plasma concentration-time profiles, area under the curve (AUC), maximum concentration (Cmax), and clearance rates. Clinical databases such as PubMed were utilized to search for studies meeting the inclusion criteria. To quantify the plasma concentration data from plasma concentration vs time plots we used the WebPlotDigitizer tool (version 4.8) [11]. The PBPK models for LTG and EFV were constructed using GastroPlus software (v.9.8.3), which enables advanced simulations of drug absorption, distribution, metabolism, and excretion by integrating detailed physiological parameters. These models were particularly enhanced by drawing upon existing literature, incorporating validated PK data and methodologies from prior studies [12, 13]. Organ weights, volumes, and blood flow rates were generated using the Population Estimates for Age-Related Physiology (PEAR Physiology™) module within GastroPlus to accurately represent the physiological state of pregnant women. To develop the LTG metabolism profile, we systematically compiled and analyzed Vmax (the maximum rate of enzymatic reaction) and Km (the substrate concentration at which the reaction rate is half of Vmax) values from existing literature, ensuring a precise definition of the metabolic parameters. For EFV, metabolism is predominantly mediated by CYP2B6, with a subsequent minor role for CYP3A4 and the involvement of UGT enzymes. The tissue distribution for EFV was modeled using a perfusion-limited approach, and tissue-to-plasma partition coefficients (Kps) were predicted using the default Lukacova model [14]. Drawing upon findings from the study conducted by Hiye Young Ji et al., we established the potential values of Vmax and Km for UGT1A1 [9]. Regarding the evaluation of the PBPK models, the predicted values for AUC and Cmax from the PBPK models were compared against observed values from clinical data. The model’s predictive performance was quantified using the percentage prediction error (%PE), where an acceptable model performance was defined as %PE values of less than 25%. Predicted plasma concentration-time profiles were plotted against observed data to visually assess the accuracy of model simulations. In addition to the conventional PK parameters, statistical metrics, including Average Fold Error (AFE), Average Absolute Fold Error (AAFE), Mean Absolute Error (MAE), and Root Mean Square Error (RMSE), were calculated. These metrics provided insights into the degree of deviation from observed values and the overall reliability of the model. At last, dynamic simulations were performed. Multiple dosing scenarios were assessed to evaluate how varying dosages impact drug exposure throughout different trimesters of pregnancy. The simulations incorporated a constant 200 mg dose of LTG in conjunction with EFV dosing adjustments based on gestational age (GA). This allowed for the quantification and comprehensive analysis of potential drug-drug interactions influenced by UGT-mediated metabolism. Results: The development and evaluation of PBPK models for LTG and EFV provided insights into their PKs during pregnancy, aligning with existing literature. The LTG model demonstrated strong predictive performance, with percentage prediction errors (PE) below 25% for AUC and Cmax, confirming its accuracy. Simulations indicated a progressive decline in LTG AUC and Cmax during pregnancy due to enhanced metabolic clearance, primarily via increased UGT1A4 enzyme activity. These findings reinforced prior research on LTG metabolism in pregnancy rather than introducing novel conclusions [15]. In contrast, the EFV PBPK model exhibited greater variability in predictive accuracy, particularly for higher doses (600 mg vs. 400 mg) and during pregnancy. Simulations suggested lower EFV AUC and Cmax in pregnant women compared to non-pregnant individuals, reflecting pregnancy-induced metabolic changes. Gender differences in EFV PKs were also noted, with males displaying higher AUC and Cmax values, likely due to genetic variations in CYP2B6 metabolism. However, these results are consistent with prior studies and do not represent new findings. Overall, while the LTG model shows potential for guiding dosing adjustments in pregnancy, further refinement is necessary for the EFV model to improve predictive reliability [15]. The innovation of our research lies in incorporating literature-derived Vmax and Km values into our models to refine the metabolic profile of LTG and EFV. This approach enabled dynamic simulations to assess potential DDIs, revealing a slight increase in LTG’s Cmax and AUC when co-administered with EFV across all gestational ages. This modest increase suggests that co-administration of EFV may facilitate a minor boost in LTG exposure, particularly in the later stages of pregnancy. However, this interaction was determined to be clinically insignificant, and adjustments to LTG dosing were not deemed necessary for maintaining therapeutic levels. The integrated analysis of enzyme activity revealed that hormonal variations and increased blood volume during pregnancy significantly impacted the metabolic clearance of LTG. Specifically, the activities of the UGT1A4 and UGT1A3 enzymes increased, contributing to the observed reduction in LTG plasma concentrations. This emphasizes the necessity of optimizing therapeutic strategies to account for these physiological changes, ensuring effective seizure management in pregnant women [15]. Conclusions: These published findings highlight the need for pregnancy-specific PBPK models. While the LTG model supports clinical use, the EFV model needs refinement. DDIs between these two drugs were not clinically relevant. This study lays the groundwork for tailored treatment in pregnant women regarding drug-drug interactions.
1. Stranix-Chibanda L, Brooks K, Eke AC. Care of Pregnant Women Living with Human Immunodeficiency Virus. Clin Perinatol. 2024 Dec;51(4):749-767. doi: 10.1016/j.clp.2024.08.010. PMID: 39487018; PMCID: PMC11724433. 2. Henderson AC, Cholli P, Lampe MA, Kourtis AP. Challenges, risks, and opportunities of antiretroviral drugs in women of reproductive potential. Expert Rev Anti Infect Ther. 2024 Apr;22(4):153-167. doi: 10.1080/14787210.2024.2334054. Epub 2024 Mar 27. PMID: 38517686; PMCID: PMC11287786. 3. Penazzato, M., Kasirye, I., Ruel, T., Mukui, I., Bekker, A., Archary, M., Musoke, P., Essajee, S., Siberry, G. K., Mahy, M., Simnoue, D., Simione, B., Zech, J. M., Mushavi, A., & Abrams, E. J. (2023). Antiretroviral postnatal prophylaxis to prevent HIV vertical transmission: Present and future strategies. Journal of the International AIDS Society, 26(2), e26032. https://doi.org/10.1002/jia2.26032 4. Gong Y, Haque S, Chowdhury P, Cory TJ, Kodidela S, Yallapu MM, Norwood JM, Kumar S. Pharmacokinetics and pharmacodynamics of cytochrome P450 inhibitors for HIV treatment. Expert Opin Drug Metab Toxicol. 2019 May;15(5):417-427. doi: 10.1080/17425255.2019.1604685. Epub 2019 Apr 20. PMID: 30951643; PMCID: PMC6497396. 5. Costa, B., Gouveia, M. J., & Vale, N. (2024). Safety and Efficacy of Antiviral Drugs and Vaccines in Pregnant Women: Insights from Physiologically Based Pharmacokinetic Modeling and Integration of Viral Infection Dynamics. Vaccines, 12(7), 782. https://doi.org/10.3390/vaccines12070782 6. Menon S, Benova L, Mabeya H. Epilepsy management in pregnant HIV+ women in sub-Saharan Africa, clinical aspects to consider: a scoping review. BMC Med. 2020 Nov 17;18(1):341. doi: 10.1186/s12916-020-01799-0. PMID: 33198766; PMCID: PMC7670685 7. Petrenaite, V., Öhman, I., Ekström, L., Sæbye, D., Hansen, T. F., Tomson, T., & Sabers, A. (2018). UGT polymorphisms and lamotrigine clearance during pregnancy. Epilepsy Research, 140, 199-208. https://doi.org/10.1016/j.eplepsyres.2018.01.011 8. Dickinson, L., Amin, J., Else, L., Boffito, M., Egan, D., Owen, A., Khoo, S., Back, D., Orrell, C., Clarke, A., Losso, M., Phanuphak, P., Carey, D., Cooper, D., & Emery, S. (2015). Pharmacokinetic and Pharmacodynamic Comparison of Once-Daily Efavirenz (400 mg vs. 600 mg) in Treatment-Naïve HIV-Infected Patients: Results of the ENCORE1 Study. Clinical Pharmacology & Therapeutics, 98(4), 406-416. https://doi.org/10.1002/cpt.156 9. Ji, H. Y., Lee, H., Lim, S. R., Kim, J. H., & Lee, H. S. (2012). Effect of Efavirenz on UDP-Glucuronosyltransferase 1A1, 1A4, 1A6, and 1A9 Activities in Human Liver Microsomes. Molecules, 17(1), 851-860. https://doi.org/10.3390/molecules17010851 10. Miners, J. O., Polasek, T. M., Hulin, J., Rowland, A., & Meech, R. (2023). Drug-drug interactions that alter the exposure of glucuronidated drugs: Scope, UDP-glucuronosyltransferase (UGT) enzyme selectivity, mechanisms (inhibition and induction), and clinical significance. Pharmacology & Therapeutics, 248, 108459. https://doi.org/10.1016/j.pharmthera.2023.108459 11. Rohatgi, A. (n.d.). WebPlotDigitizer (Version [4.8]). https://automeris.io/WebPlotDigitizer 12. Caleffi-Marchesini, E.R.; Borghi-Pangoni, F.B.; Macente, J.; Chiamulera-Mantovani, P.; Mazucheli, J.; Cristofoletti, R.; Diniz, A. Exploring in Vitro Solubility of Lamotrigine in Physiologically Mimetic Conditions to Prospect the in Vivo Dissolution in Pediatric Population. Biopharm. Drug Dispos. 2023, 44, 147–156 13. Simulation Plus Efavirenz Physiologically Based Pharmacokinetic Model Development and Validation as a Moderate CYP3A4 Inducer for Drug-Drug Interaction Predictions. Available online: https://www.simulations-plus.com/wp-content/uploads/MIDD-Inger-Darling-efavirenz.pdf 14. Lukacova, P.N., Lave, T., Fraczkiewicz, G., Bolger, M.B. & Woltosz, W.S. General approach to calculation of tissue:plasma partition coefficients for physiologically based pharmacokinetic (PBPK) modeling, Poster Session presented at: 2008 AAPS National Annual Meeting and Exposition, November 16-20, 2008, Atlanta, GA. 15. Costa, B., Gouveia, M. J., & Vale, N. (2024). PBPK Modeling of Lamotrigine and Efavirenz during Pregnancy: Implications for Personalized Dosing and Drug-Drug Interaction Management. Pharmaceutics, 16(9), 1163. https://doi.org/10.3390/pharmaceutics16091163
Reference: PAGE 33 (2025) Abstr 11382 [www.page-meeting.org/?abstract=11382]
Poster: Drug/Disease Modelling - Absorption & PBPK